trainer_state.json

{
  "best_metric": 0.9197333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0920/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": Infinity,
      "learning_rate": 8.580008580008581e-05,
      "loss": 0.7634,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8488,
      "eval_loss": 0.4925871193408966,
      "eval_runtime": 10.1108,
      "eval_samples_per_second": 370.891,
      "eval_steps_per_second": 5.835,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.245128154754639,
      "learning_rate": 7.151437151437152e-05,
      "loss": 0.248,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.852,
      "eval_loss": 0.5061822533607483,
      "eval_runtime": 10.3962,
      "eval_samples_per_second": 360.708,
      "eval_steps_per_second": 5.675,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.0250239335000515,
      "learning_rate": 5.7228657228657234e-05,
      "loss": 0.1198,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8712,
      "eval_loss": 0.48067396879196167,
      "eval_runtime": 10.0924,
      "eval_samples_per_second": 371.566,
      "eval_steps_per_second": 5.846,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.01381666585803032,
      "learning_rate": 4.294294294294294e-05,
      "loss": 0.0589,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8888,
      "eval_loss": 0.43756797909736633,
      "eval_runtime": 10.0064,
      "eval_samples_per_second": 374.76,
      "eval_steps_per_second": 5.896,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0002999190182890743,
      "learning_rate": 2.8657228657228656e-05,
      "loss": 0.0258,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9029333333333334,
      "eval_loss": 0.4034755527973175,
      "eval_runtime": 10.387,
      "eval_samples_per_second": 361.03,
      "eval_steps_per_second": 5.68,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0002787703706417233,
      "learning_rate": 1.4371514371514371e-05,
      "loss": 0.0075,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9130666666666667,
      "eval_loss": 0.37668418884277344,
      "eval_runtime": 10.3004,
      "eval_samples_per_second": 364.063,
      "eval_steps_per_second": 5.728,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.002461371012032032,
      "learning_rate": 8.580008580008581e-08,
      "loss": 0.0016,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9197333333333333,
      "eval_loss": 0.3552483916282654,
      "eval_runtime": 10.1999,
      "eval_samples_per_second": 367.652,
      "eval_steps_per_second": 5.784,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}